Directly imaging a terrestrial planet is going to be a tough challenge. Suppose you were thirty light years from the Sun, looking back at our star in the hope of seeing the Earth. You would face the problem that the Earth and its star show an angular separation of 100 milliarcseconds, a spacing so tiny that the far brighter Sun would render its third planet (and all the others) invisible. Indeed, in optical wavelengths the Earth is ten billion times less bright than the Sun. How to go about seeing it?

Observing at other wavelengths offers some help. The Sun is only a million times brighter than the Earth in the mid-infrared, which is why our first glimpse of planets like ours will probably be in this range.

And it may be that our first catch is not a mature, established planet potentially offering a habitat to living organisms. Instead, it may be a clump of molten rock still glowing brightly from the heat of formation. Even after surface magma solidifies — and new work suggests this could take five million years, rather than the hundreds of thousands previously thought — the planet might stay hot enough to be an unusually bright target in the infrared for tens of millions more.

This is the conclusion of Linda Elkins-Tanton (MIT), whose work implies that a glowing, molten planetary surface may be the most feasible find for early terrestrial planet hunter missions. As to the processes producing all that magma, they’re initially the result of radioactivity in the planet’s interior and the heat of planetary formation created by millions of rocky collisions in the early system. But a second process, causing iron-rich materials to sink toward the core, may force hotter materials from within back up to the surface, keeping the landscape molten much longer.

So we may have a ‘magma ocean’ that lasts at least a few million years longer than had previously been thought. It’s an interesting model, and one that clearly has implications for detectability since it lengthens the window for observation. Surprisingly, the theory may gain support when the MESSENGER mission settles in around Mercury. Earth’s crust is too dynamic for material from such early epochs to survive, but Mercury’s surface may offer up minerals that Elkins-Tanton’s model says should be there. Even better, of course, would be the direct detection of a molten, young Earth analog, but for now Mercury may have to do.